much better git related version numbering in our (*nix for now) build process

The net result of this patch is to make available via Config.pm and -v/-V the
details about the git version info we have available for the build. When built within
a git repository git is queried directly. When built from a snapshot or bundle
it is assumed that the source is unchanged, and that the required details are
avaialble in a file called .patch, whose format current is a four field string
in the following format: "$branchname $date.$time $sha1 $describe". The
generator of these files currently resides on camel.booking.com.

* git-describe is now used more directly with -v.
	When the prefix of git-describe matches the version number
	as determined by the defines in patchlevel.h then we use ONLY
	the git-describe output, otherwise we include
	the git describe in parenthesis after the version number. Either way
	the describe text is optionally followed by a star should there be
	uncommitted changes.
	eg: This is perl, v5.11.0 (GitLive-blead-136-g58ca560) built for i686-linux
	or: This is perl, v5.11.0-1-g58ca560 built for i686-linux
	or: This is perl, v5.11.0 built for i686-linux

* include the SHA1 in perl -V summary, and automatically include unpushed
	commits in the registered patches list

* include various git/version/.patch details in %Config, as follows:

	git_commit_id             # sha1 of HEAD
	git_ancestor              # ancestor in $remote/$branch (presumably canonical)
	git_describe              # git describe
	git_branch                # current branch
	git_uncommitted_changes   # "true" if there are any, empty otherwise
	git_unpushed_commits      # List of sha1's of unpushed commits
	git_commit_id_title       # Used to make the perl -V summary output

Additionally one more value is added depending on build process used: when
building from an rsynced snapshot (or any dist including a file called
.patch) then the second field will be used to populate the
"git_snapshot_date" field. Otherwise if built in a git directory (as is hopefully
recommended these day) then the field will be "git_commit_date" which will be the
commit date of HEAD.

This patch introduces two new files (on top of .patchnum) that will be generated by
make_patchnum.sh: "lib/Config_git.pl" and "unpushed.h", the former is used to make
git data available to Config.pm/%Config without rebuilding everything else, and the
second is used to expose unpushed commits (if any) via the registered patch facility
of patchlevel.h

12 files changed, 149 insertions, 54 deletions

diff --git a/.gitignore b/.gitignore
index 4b430f9485..362fc7a319 100644
--- a/.gitignore
+++ b/.gitignore
@@ -5,6 +5,7 @@
 # these are generated by make_patchnum.sh from git or from a .patchfile
 .patchnum
 .sha1
+unpushed.h
 
 # ignore bug*.pl
 bug*.pl
diff --git a/Makefile.SH b/Makefile.SH
index ff50cfd0d7..9ad8b6f102 100644
--- a/Makefile.SH
+++ b/Makefile.SH
@@ -371,7 +371,10 @@ CCCMD    = `sh $(shellflags) cflags "optimize='$(OPTIMIZE)'" $@`
 
 CCCMDSRC = `sh $(shellflags) cflags "optimize='$(OPTIMIZE)'" $<`
 
+# we dont include lib/Config_git.pl here, as it causes circular dependencies
 CONFIGPM = lib/Config.pm lib/Config_heavy.pl
+# so we add it here:
+CONFIGPM_EXTRA = $(CONFIGPM) lib/Config_git.pl
 
 CONFIGPOD = lib/Config.pod
 
@@ -540,7 +543,8 @@ sperl.i: perl.c $(h)
 make_patchnum:
 	sh $(shellflags) make_patchnum.sh
 
-perl$(OBJ_EXT): .patchnum .sha1
+# make sure that we recompile perl.c if .patchnum changes
+perl$(OBJ_EXT): .patchnum unpushed.h
 
 translators:	miniperl$(EXE_EXT) $(CONFIGPM) FORCE
 	@echo " "; echo "	Making x2p stuff"; cd x2p; $(LDLIBPTH) $(MAKE) all
@@ -1217,7 +1221,8 @@ _tidy:
 	-cd utils; $(LDLIBPTH) $(MAKE) clean
 	-cd x2p; $(LDLIBPTH) $(MAKE) clean
 	-rm .patchnum
-	-rm .sha1
+	-rm lib/Config_git.pl
+	-rm unpushed.h
 	-@for x in $(DYNALOADER) $(dynamic_ext) $(static_ext) $(nonxs_ext) ; do \
 	$(LDLIBPTH) sh ext/util/make_ext clean $$x MAKE=$(MAKE) ; \
 	done
diff --git a/cflags.SH b/cflags.SH
index 6e6e8a950e..132f466a92 100755
--- a/cflags.SH
+++ b/cflags.SH
@@ -297,12 +297,12 @@ for file do
     opmini) ;;
     pad) ;;
     perl) 
-        if [ -f .patchnum -a -n "$(cat .patchnum)" ] ; then
-                ccflags="-DPERL_PATCHNUM=`cat .patchnum` $ccflags"
-        fi
-        if [ -f .sha1 -a -n "$(cat .sha1)" ] ; then
-                ccflags="-DPERL_GIT_SHA1=`cat .sha1` $ccflags"
+        if [ -f .patchnum -a -n "$(awk 'BEGIN{ORS=""} /describe:/ {print $1}' .patchnum)" ] ; then
+                ccflags="-DPERL_PATCHNUM=$(awk 'BEGIN{ORS=""} /describe:/ {print $2}' .patchnum) $ccflags"
         fi
+	if [ -f .patchnum -a -n "$(awk 'BEGIN{ORS=""} /status:/ {print $2}' .patchnum)" ] ; then
+                ccflags="-DPERL_GIT_UNCOMMITTED_CHANGES=$(awk 'BEGIN{ORS=""} /status:/ {print $2}' .patchnum) $ccflags"
+	fi
         ;;
     perlapi) ;;
     perlmain) ;;
diff --git a/configpm b/configpm
index d781b50c6c..af6f65a0c8 100755
--- a/configpm
+++ b/configpm
@@ -458,7 +458,19 @@ my $summary_expanded;
 sub myconfig {
     return $summary_expanded if $summary_expanded;
     ($summary_expanded = $summary) =~ s{\$(\w+)}
-		 { my $c = $Config::Config{$1}; defined($c) ? $c : 'undef' }ge;
+		 { 
+			my $c;
+			if ($1 eq 'git_ancestor_line') {
+				if ($Config::Config{git_ancestor}) {
+					$c= "\n  Ancestor: $Config::Config{git_ancestor}";
+				} else {
+					$c= "";
+				}
+			} else {
+                     		$c = $Config::Config{$1}; 
+			}
+			defined($c) ? $c : 'undef' 
+		}ge;
     $summary_expanded;
 }
 
@@ -535,6 +547,11 @@ foreach my $prefix (qw(libs libswanted)) {
 
 $heavy_txt .= "EOVIRTUAL\n";
 
+$heavy_txt .= <<'ENDOFGIT';
+require 'Config_git.pl';
+$Config_SH_expanded .= $Config::Git_Data;
+ENDOFGIT
+
 $heavy_txt .= $fetch_string;
 
 $config_txt .= <<'ENDOFEND';
diff --git a/install_lib.pl b/install_lib.pl
index 7eeae1d33c..ae8ba0a990 100644
--- a/install_lib.pl
+++ b/install_lib.pl
@@ -29,6 +29,7 @@ OS
 	undef %Config::;
 	delete $INC{"Config.pm"};
 	delete $INC{"Config_heavy.pl"};
+	delete $INC{"Config_git.pl"};
 	# You never saw us. We weren't here.
 
 	require Config;
diff --git a/installperl b/installperl
index da8f45ee4a..d8d36be7d6 100755
--- a/installperl
+++ b/installperl
@@ -759,7 +759,7 @@ sub installlib {
     if ($dir =~ /^auto/ ||
 	  ($name =~ /^(.*)\.(?:pm|pod)$/ && $archpms{$1}) ||
 	  ($name =~ /^(.*)\.(?:h|lib)$/i && ($Is_W32 || $Is_NetWare)) ||
-	  $name eq 'Config_heavy.pl'
+	  $name=~/^Config_(heavy\|git)\.pl\z/
        ) {
 	$installlib = $installarchlib;
 	return unless $do_installarchlib;
diff --git a/lib/.gitignore b/lib/.gitignore
index e6facd30ef..0fa004e77b 100644
--- a/lib/.gitignore
+++ b/lib/.gitignore
@@ -4,6 +4,7 @@
 /Config.pm
 /Config.pod
 /Config_heavy.pl
+/Config_git.pl
 /DB_File.pm
 /Data
 /Devel/DProf.pm
diff --git a/make_patchnum.sh b/make_patchnum.sh
index 9a2f946b18..c9c74af683 100644
--- a/make_patchnum.sh
+++ b/make_patchnum.sh
@@ -1,35 +1,85 @@
 #!/bin/sh
 
-. ./config.sh
+# this script is used to regenerate a number of special build files
+# based on either information contained in a file called .patch or
+# directly from git.
+# The files involved are:
+#   .patchnum         # information about the current checkout
+#   lib/Config_git.pl # holds some special configure settings related to git
+#   unpushed.h        # header file used by patchlevel.h to store unpushed commits
 
-Existing=`cat .patchnum 2>/dev/null`
-Existing_Sha1=`cat .sha1 2>/dev/null`
+existing_patchnum=$(cat .patchnum 2>/dev/null)
+existing_config=$(cat lib/Config_git.pl 2>/dev/null)
+existing_unpushed=$(cat unpushed.h 2>/dev/null)
 
-if test -s ".patch" ; then
-	Current=`awk '{print $4}' .patch`
-	Sha1=`awk '{print $3}' .patch`
-elif test -d ".git" ; then
-	# we should do something better here
-	Current=`git describe`
-	Sha1=`git rev-parse HEAD`
-	Changed=`git diff-index --name-only HEAD`
-	test -n "$Changed"  && Current="$Current-with-uncommitted-changes"
-fi
+unpushed_commits='/*no-op*/'
+if [ -s ".patch" ] ; then
+	read branch snapshot_created commit_id describe < .patch
+	changed=""
+	extra_info="git_snapshot_date='$snapshot_created'"
+	commit_title='Snapshot of:'
+elif [ -d ".git" ]; then
+	branch=$(git branch | awk 'BEGIN{ORS=""} /\*/ { print $2 }')
+	test -n "$branch" && remote=$(git config branch.$branch.remote)
+	commit_id=$(git rev-parse HEAD)
+	changed=$(git diff-index --name-only HEAD)
+	describe=$(git describe --tags)
+	commit_created=$(git log -1 --pretty='format:%ci')
+	new_patchnum="describe: $describe"
+	extra_info="git_commit_date='$commit_created'"
+	if [ -n "$branch" ] && [ -n "$remote" ]; then
+		unpushed_commit_list=$(git cherry $remote/$branch | awk 'BEGIN{ORS=","} /+/ {print $2}' | sed -e 's/,$//')
+		unpushed_commits=$(git cherry $remote/$branch | awk 'BEGIN{ORS="\t\\\n"} /+/ {print ",\"" $2 "\""}')
 
-if test "$Existing" != "$Current" -o "$Existing_Sha1" != "$Sha1" ; then
-	(echo "hi there\c" ; echo " ") >.echotmp
-	if $contains c .echotmp >/dev/null 2>&1 ; then
-		n='-n'
-		c=''
-	else
-		n=''
-		c='\c'
+		if [ -n "$unpushed_commits" ]; then
+			commit_title="Local Commit:"
+			ancestor=`git rev-parse $remote/$branch`
+			extra_info="$extra_info
+git_ancestor='$ancestor'
+git_unpushed='$unpushed_commit_list'"
+		fi
+			
+	fi
+	if [ -n "$changed" ]; then
+		changed="true"
+		commit_title="Derived from:"
+		new_patchnum="$new_patchnum
+status: uncommitted-changes"
 	fi
-	rm -f .echotmp
-	echo "Updating .patchnum and .sha1"
-	echo $n "$Current$c" > .patchnum
-	echo $n "$Sha1$c" > .sha1
+	test -z "$commit_title" && commit_title='Commit id:'
+fi
 
+new_unpushed=$(cat <<EOFTEXT
+/*********************************************************************
+* WARNING: unpushed.h is automatically generated by make_patchnum.sh *
+*          DO NOT EDIT DIRECTLY - edit make_patchnum.sh instead      *
+*********************************************************************/
+#define PERL_GIT_UNPUSHED_COMMITS       $unpushed_commits
+/*leave-this-comment*/
+EOFTEXT
+)
+new_config=$(cat <<EOFDATA
+#################################################################
+# WARNING: lib/Config_git.pl is generated by make_patchnum.sh   #
+#          DO NOT EDIT DIRECTLY - edit make_patchnum.sh instead #
+#################################################################
+\$Config::Git_Data=<<'ENDOFGIT';
+git_commit_id='$commit_id'
+git_describe='$describe'
+git_branch='$branch'
+git_uncommitted_changes='$changed'
+git_commit_id_title='$commit_title'
+$extra_info
+ENDOFGIT
+EOFDATA
+)
+# only update the files if necessary, other build product depends on these files
+if [ "$existing_patchnum" != "$new_patchnum" ] || [ "$new_config" != "$existing_config" ] || [ "$existing_unpushed" != "$new_unpushed" ]; then
+	echo "Updating .patchnum and lib/Config_git.pl"
+	echo "$new_patchnum" > .patchnum
+	echo "$new_config" > lib/Config_git.pl
+	echo "$new_unpushed" > unpushed.h
 else
-	echo "Reusing .patchnum and .sha1" 
+	echo "Reusing .patchnum and lib/Config_git.pl"
 fi
+
diff --git a/myconfig.SH b/myconfig.SH
index 3dfd9537c5..f5b432ad31 100644
--- a/myconfig.SH
+++ b/myconfig.SH
@@ -28,6 +28,7 @@ $startsh
 # Note that the text lines /^Summary of/ .. /^\s*$/ are copied into Config.pm.
 cat <<'!NO!SUBS!'
 Summary of my $package (revision $revision $version_patchlevel_string) configuration:
+  $git_commit_id_title $git_commit_id$git_ancestor_line
   Platform:
     osname=$osname, osvers=$osvers, archname=$archname
     uname='$myuname'
diff --git a/patchlevel.h b/patchlevel.h
index 391cb4126c..a17a6e679d 100644
--- a/patchlevel.h
+++ b/patchlevel.h
@@ -116,9 +116,21 @@ hunk.
  */
 
 #if !defined(PERL_PATCHLEVEL_H_IMPLICIT) && !defined(LOCAL_PATCH_COUNT)
+#include "unpushed.h"
+#if !defined(PERL_GIT_UNPUSHED_COMMITS)
+#define PERL_GIT_UNPUSHED_COMMITS_PATCH /* no-op */
+#else
+#define PERL_GIT_UNPUSHED_COMMITS_PATCH PERL_GIT_UNPUSHED_COMMITS
+#endif
+#if !defined(PERL_GIT_UNCOMMITTED_CHANGES)
+#define PERL_GIT_UNCOMMITTED_CHANGES_PATCH /*no op*/
+#else
+#define PERL_GIT_UNCOMMITTED_CHANGES_PATCH , STRINGIFY(PERL_GIT_UNCOMMITTED_CHANGES)
+#endif
 static const char * const local_patches[] = {
 	NULL
-	,"DEVEL"
+	PERL_GIT_UNPUSHED_COMMITS_PATCH    /* do not remove this line */
+        PERL_GIT_UNCOMMITTED_CHANGES_PATCH /* do not remove this line */
 	,NULL
 };
 
diff --git a/perl.c b/perl.c
index 7c5da008ad..d224ca6d1d 100644
--- a/perl.c
+++ b/perl.c
@@ -1904,7 +1904,6 @@ S_parse_body(pTHX_ char **env, XSINIT_t xsinit)
 #else
 		    sv_catpvs(opts_prog,"\"\\nCharacteristics of this binary (from libperl): \\n");
 #endif
-		    sv_catpvs(opts_prog,"  Source revision: " STRINGIFY(PERL_GIT_SHA1) "\\n");
 		    sv_catpvs(opts_prog,"  Compile-time options: $_\\n\",");
 
 #if defined(LOCAL_PATCH_COUNT)
@@ -1914,16 +1913,8 @@ S_parse_body(pTHX_ char **env, XSINIT_t xsinit)
 				 "\"  Locally applied patches:\\n\",");
 			for (i = 1; i <= LOCAL_PATCH_COUNT; i++) {
 			    if (PL_localpatches[i])
-#ifdef X_PERL_PATCHNUM
-/* this is ifdef'ed out, we would enable this if we want to transform 
-   "DEVEL" registered patches into the git name */
-				if (strEQ(PL_localpatches[i],"DEVEL"))
-				    Perl_sv_catpvf(aTHX_ opts_prog,"q%c\t%s\n%c,",
-					       0, STRINGIFY(PERL_PATCHNUM), 0);
-				else
-#endif
-				    Perl_sv_catpvf(aTHX_ opts_prog,"q%c\t%s\n%c,",
-					       0, PL_localpatches[i], 0);
+				Perl_sv_catpvf(aTHX_ opts_prog,"q%c\t%s\n%c,",
+				    0, PL_localpatches[i], 0);
 			}
 		    }
 #endif
@@ -3315,11 +3306,29 @@ Perl_moreswitches(pTHX_ const char *s)
 	if (!sv_derived_from(PL_patchlevel, "version"))
 	    upg_version(PL_patchlevel, TRUE);
 #if !defined(DGUX)
-	PerlIO_printf(PerlIO_stdout(),
+	{
+	    SV* level= vstringify(PL_patchlevel);
+#ifdef PERL_PATCHNUM
+	    SV* num= newSVpvn(STRINGIFY(PERL_PATCHNUM),sizeof(STRINGIFY(PERL_PATCHNUM))-1);
+#ifdef PERL_GIT_UNCOMMITTED_CHANGES
+	    sv_catpvf(num,"%s","*");
+#endif
+
+	    if (sv_len(num)>=sv_len(level) && strnEQ(SvPV_nolen(num),SvPV_nolen(level),sv_len(level))) {
+		SvREFCNT_dec(level);
+		level= num;
+	    } else {
+		sv_catpvf(level, " (%s)",SvPV_nolen(num));
+		SvREFCNT_dec(num);
+	    }
+ #endif
+	    PerlIO_printf(PerlIO_stdout(),
 		"\nThis is perl, %"SVf
 		" built for %s",
-		SVfARG(vstringify(PL_patchlevel)),
+		level,
 		ARCHNAME);
+	    SvREFCNT_dec(level);
+	}
 #else /* DGUX */
 /* Adjust verbose output as in the perl that ships with the DG/UX OS from EMC */
 	PerlIO_printf(PerlIO_stdout(),
@@ -3332,9 +3341,6 @@ Perl_moreswitches(pTHX_ const char *s)
 			Perl_form(aTHX_ "        OS Specific Release: %s\n",
 					OSVERS));
 #endif /* !DGUX */
-#if defined PERL_PATCHNUM
-	PerlIO_printf(PerlIO_stdout(),"\nCompiled from: %s",STRINGIFY(PERL_PATCHNUM));
-#endif
 #if defined(LOCAL_PATCH_COUNT)
 	if (LOCAL_PATCH_COUNT > 0)
 	    PerlIO_printf(PerlIO_stdout(),
diff --git a/t/run/switches.t b/t/run/switches.t
index 0d91d8a810..a9e23ea196 100644
--- a/t/run/switches.t
+++ b/t/run/switches.t
@@ -260,10 +260,11 @@ SWTESTPM
 
 {
     local $TODO = '';   # these ones should work on VMS
-
+    # there are definitely known build configs where this test will fail
+    # DG/UX comes to mind. Maybe we should remove these special cases?
     my $v = sprintf "%vd", $^V;
     like( runperl( switches => ['-v'] ),
-	  qr/This is perl, v$v (?:DEVEL:[-A-Za-z0-9]+ )?built for \Q$Config{archname}\E.+Copyright.+Larry Wall.+Artistic License.+GNU General Public License/s,
+	  qr/This is perl, v$v(?:[-\w]+| \([^)]+\))? built for \Q$Config{archname}\E.+Copyright.+Larry Wall.+Artistic License.+GNU General Public License/s,
           '-v looks okay' );
 
 }